NCRC Proposal for Underserved Tracts Would Increase Lending in Communities of Color by Billions of Dollars

Executive Summary

  • NCRC proposes to add underserved census tracts as a criterion on component tests on CRA exams. NCRC identifies underserved tracts as those that receive low levels of lending per housing units and small businesses.
  • A significant portion of these tracts are predominantly minority.
  • NCRC estimated that adding these tracts on CRA exams would increase retail lending in communities of color by more than $10 billion over five years.
  • The NCRC proposal would effectively address regional and racial inequalities by targeting counties with significant African American and Hispanic populations in the South, Midwest and West.
  • The OCC final rule added underserved communities but its definition is different and its final rule did not indicate if it targeted the most underserved communities.


The Community Reinvestment Act (CRA) is an income-based law. Accordingly, the regulations and exams focus on evaluating bank lending, investing and services to low- and moderate-income (LMI) borrowers and tracts. However, one of the major motivations prompting the passage of CRA in 1977 was reversing redlining and disinvestment from communities of color as well as working class neighborhoods. Community advocates and local public sector officials testified during the hearings before the vote on CRA that redlining was pervasive in communities of color.[1]

Over the years, the National Community Reinvestment Coalition (NCRC) and our members have advocated for the inclusion of race on CRA exams since people of color and communities of color continue to experience a shortage of affordably priced banking products and a disproportionate amount of high-cost and abusive products. The financial crisis of 2008 stemmed in large part from abusive lenders targeting communities of color that were starved of mainstream banking products.[2] Currently, communities of color still encounter shortages of loans.

The regulatory agencies have opted not to include communities of color explicitly on CRA exams. However, the Office of the Comptroller of the Currency (OCC)’s recent final rule on CRA added a criterion of underserved tracts that applies to urban as well as rural counties. Although NCRC disagrees with the methodology the OCC used to identify these tracts, the OCC’s final rule recognized the issue of underserved communities. Therefore, NCRC anticipates that the agencies (particularly the Federal Reserve Board and FDIC that have not proposed changes to CRA) might be amenable to applying the concept of underserved tracts as a means of focusing more attention on underserved communities, including communities of color.

In a previous report, NCRC developed a methodology for identifying census tracts that were underserved as measured by receiving low levels of retail lending per housing units and small businesses.[3] This report divided census tracts into quintiles based on lending levels per housing units and small businesses. The average percentage of people of color in Quintile 1 tracts (the lowest quintile in terms of loans) was 56.8%.

NCRC recommended that the tracts constituting Quintile 1 tracts be added to the component tests of CRA exams. For example, the retail lending test of the large bank exam currently assesses the percentage of loans in low- and moderate-income (LMI) tracts as well as to LMI borrowers. This critical evaluation of lending in LMI tracts would remain in the NCRC proposal. Under the NCRC proposal, the lending test would also assess the percentage of loans in underserved tracts. In terms of lending in underserved tracts, banks would be compared to each other and to demographic benchmarks such as the percentage of owner occupied housing units in underserved tracts. Including underserved tracts on CRA exams would motivate banks to increase their lending in these tracts.

It is not possible to estimate precisely how much retail lending would increase in underserved tracts. However, research conducted by economists at the Federal Reserve Bank of Philadelphia provided some hints as to how much lending could increase. This research found that removing tracts from CRA eligibility would decrease home lending by 10% to 20%.[4] This exercise would be the reverse: adding a group of tracts to CRA eligibility. Since the previous research did not model how adding tracts as CRA eligible would impact lending, NCRC will assume that the increase in lending is on the low end, that is a 10% increase.

An additional twist in estimating the impact is identifying tracts in Quintile 1 that are not only predominantly minority (more than 50% of the residents are people of color) but are also middle- and upper-income. LMI tracts in Quintile 1 already benefit from CRA. The tracts that would be newly eligible for CRA would be middle- and upper-income tracts. As discussed above, people of color and communities of color at all income levels have been persistently undeserved by banks. Thus, it is desirable for CRA to identify non-LMI communities of color that are underserved and to encourage banks to make more loans in them. Thus, the increases in lending discussed immediately below are those in middle- and upper-income communities of color. There could also be lending increases in LMI tracts in Quintile 1 but this paper errs on the side of a conserative estimate by not attempting to predict increases in lending in the LMI tracts.

As mentioned in the previous paper, NCRC recommended that the agencies would update the  list of underserved tracts periodically. This would result in tracts being removed from Quintile 1 if they had been successfully revitalized with higher lending levels. New tracts would be added to Quintile 1 based on updated lending and census data that indicated that these tracts had low levels of lending. Thus, NCRC’s proposal for considering underserved tracts is dynamic and would focus attention only on those non-LMI tracts that were in need of more lending as revealed by the most current data.


The tables display lending increases at a county level for groups of newly CRA eligible census tracts. Overall, 1,265 predominantly minority tracts in 344 counties would be newly eligible for CRA as shown in Table 1. NCRC calculated small business and home lending over the last five years in these tracts. A 10% increase in this lending would result in more than $10 billion over five years in home and small business lending in these tracts.

Three states stand out as significant beneficiaries as shown in Table 2. California’s communities of color would gain more than $4 billion in retail lending, New York would gain $1.1 billion and Florida would gain almost $1 billion. In California, counties with large increases in several census tracts in lending include Los Angeles, Alameda and San Diego. In New York, Queens and Manhattan have large dollar increases in lending. In Florida, the counties with significant new lending include Miami, Hillsborough, Orange and Broward. Texas, Virginia, New Jersey, Illinois, Georgia, Hawaii and Arizona round out the top ten states in terms of the dollar amount of new lending. Some of the large dollar gains by state is due to the expensive housing stock. For example, a number of California counties have per capita increases in lending ranging from $10,000 to $17,000.

A map of the regional distribution of counties makes it clear that the beneficiaries of NCRC’s proposal would not be just expensive coastal areas. The greatest number of counties are in the South and Southwest with some clusters in the Midwest. This finding is significant in that a Brookings paper states that most African Americans remain in the South where poverty and distressed economic conditions afflict them.[5] In addition, the California counties have large Hispanic populations with a significant segment experiencing economic disadvantages; counties in the South also exhibit these characteristics.[6] Thus, NCRC’s proposal to add underserved areas is most likely addressing regional and racial inequalities.

Download Lending Trends and Increases by State and County table and Lending Increase by State table here.

Lending Increase by State


NCRC identified census tracts that were underserved in terms of retail loans per housing unit and small business. These tracts contained a high number of communities of color. NCRC then isolated a subset of tracts among the underserved tracts that were predominantly minority and were not LMI. Under NCRC’s CRA reform proposal, these tracts would be included in the component tests of CRA exams. Including them in the retail lending test would increase lending over five years by about $10 billion. This proposal would also likely address regional racial inequalities directly. In contrast to our proposal, the OCC proposed to add underserved tracts to CRA exams but did not test what tracts and what regions of the country would benefit from its proposal. The OCC approach is ill equipped to reform CRA in a manner that benefits underserved communities.

[1] For a review of the CRA hearings and passage of CRA in 1977, see Josh Silver, The Purpose And Design Of The Community Reinvestment Act (CRA): An Examination Of The 1977 Hearings And Passage Of The CRA, https://ncrc.org/the-purpose-and-design-of-the-community-reinvestment-act-cra-an-examination-of-the-1977-hearings-and-passage-of-the-cra/

[2] For an overview of continued racial disparities in lending, see Aaron Glantz and Emmanuel Martinez, For People of Color, Banks are Shutting the Door on Homeownership, February 15, 2018, https://www.revealnews.org/article/for-people-of-color-banks-are-shutting-the-door-to-homeownership//. For documentation of high cost lending targeting communities of color, see NCRC, Foreclosure in the Nation’s Capital: How Unfair and Reckless Lending Undermines Homeownership, April 2010, https://ncrc.org/foreclosure-in-the-nations-capital-how-unfair-and-reckless-lending-undermines-homeownership/. Also, see NCRC, Income is No Shield Against Racial Disparities in Lending II: A Comparison of High-Cost Lending In America’s Metropolitan and Rural Areas, July 2008,  https://ncrc.org/wp-content/uploads/2008/07/income%20is%20no%20shield%20ii.pdf

[3] Bruce Mitchell, PhD. and Josh Silver, Adding Underserved Census Tracts As Criterion On CRA Exams, January 2020, NCRC, https://ncrc.org/adding-underserved-census-tracts-as-criterion-on-cra-exams/

[4] Lei Ding and Leonard Nakamura, Don’t Know What You Got Till It’s Gone: The Effects of the Community Reinvestment Act (CRA) on Mortgage Lending in the Philadelphia Market, Working Paper No. 17-15, June 19, 2017, https://www.philadelphiafed.org/-/media/research-and-data/publications/working-papers/2017/wp17-15.pdf

[5]Jay Shambaugh, Ryan Nunn, and Stacy A. Anderson, How racial and regional inequality affect economic opportunity, Brookings, February 2019, https://www.brookings.edu/blog/up-front/2019/02/15/how-racial-and-regional-inequality-affect-economic-opportunity/

[6] Economic inequality: The worst states for Hispanics and Latinos, Horus News Now, January 2018,  https://horusnewsnow.com/economic-inequality-the-worst-states-for-hispanics-and-latinos/. Also see this map from the Census Bureau webpage. Percent Below Poverty Level—Population For Whom Poverty Status Is Determined—RACE AND HISPANIC OR LATINO ORIGIN—Hispanic Or Latino Origin (Of Any Race)—2018 ACS Five Year Estimate at a county level.


Josh Silver, Senior Advisor, Policy 
Jason Richardson, Director, Research & Evaluation

Print Friendly, PDF & Email
Scroll to Top